自组织特征映射神经网络在说话人识别中的应用

The Application of SOM Neural Network In Speaker Recognition

Mel频率倒谱系数MFCC在说话人识别中已被广泛使用。本文以MFCC为声音信号的特征参数提取的标准。识别部分采用自组织神经网络的方法进行建模,实验结果表明该方法有较高的识别率。

Mel Frequency Cepstral Coefficients in the MFCC Speaker Recognition has been widely used。 In this paper,the voice signal to the MFCC feature extraction. Used to identify some of the self-organizing neural network approach to model 。 The results show that the method with a higher recognition rate.